Breast cancer is the most commonly diagnosed cancer among women in the United States. There is currently a controversy on whether the Epstein-Barr virus (EBV; Human Herpes Virus 4) plays a role in breast carcinogenesis. We propose to test the hypothesis that EBV is commonly found in invasive ductal carcinomas of the breast but not in normal adjacent tissue, and that the presence of EBV is associated with a more aggressive phenotype, a poor prognosis, and epidemiolgic risk factors that are associated with suppression of the cellular immune system including a diet high in animal fat or animal protein, a history of blood transfusion, and cigarette smoking. This study will be conducted in the Iowa Women's Health Study (IWHS), a prospective cohort study of postmenopausal breast cancer risk in over 40,000 older Iowa women who were first enrolled in 1986. We propose to retrieve tumor blocks from incident invasive ductal carcinomas first diagnosed in the IWHS between 1988 and 1990 (N=258) in collaboration with the State Health Registry of Iowa. Based on extensive prior experience, we expected to have 184 useable tumor blocks available for analysis. Diagnostic slides and pathology reports will be reviewed in order to grade the tumors and to select appropriate normal and tumor sections from paraffin blocks. Tumor blocks will be processed in th Mayo Tissue Acquisition and Processing Core. We will attempt to identify the presence of EBV in the tumor and adjacent histologically normal tissue using PCR-based methods to identify EBV, and immunohistochemical methods to identify EBNA-1 expression, allowing us to estimated the prevalence of EBV (any EBV; EBV-1; EBV-2) infection in a relatively large sample of invasive ductal carcinomas. We will also correlate EBV status in tumors with prognostic factors including tumor size, axillary lymph node status, estrogen and progesterone receptor status, and histologic grade. As secondary aims, we will evaluate whether EBV status predicts long-term (great than or equal to 10-year) survival after breast cancer diagnosis (overall survival; death due to breast cacncer), and whether certain epidemiologic risk factors that alter immune function allowing reactivation of EBV infection are more strongly associated with EBV+breast cancers. We have carefully designed this study within the resource constraints of the small grant program in order to clearly address our main hypotheses, and to begin to address our secondary aims, albeit with limited power, in order to generate preliminary data on which to base future studies within this cohort as well as other study populations. This proposal also fits well into the goals of the small grants program in cancer epidemiology since it address an emerging, high risk/high gain hypothesis that requires the types of data we will generate in order to justify a more definitive study.